Driveseg: dynamic driving scene segmentation data set

Click on the above “ Xiaobai studies vision ”, Optional plus " Star standard " or “ Roof placement ”

 Heavy dry goods , First time delivery 

【 Reading guide 】 MIT and Toyota released DriveSeg Data set to accelerate autopilot research ,DriveSeg Contains precise pixel level representations of many common road objects , And through the lens of continuous video driving scene .

How do we train autopilot models , To deepen our understanding of the world around us ？ Can computers learn from past experience to recognize future patterns , To help them safely cope with new unpredictable situations ？

MIT transportation and logistics center AgeLab Cooperate with Toyota Safety Research Center （CSRC） Of the researchers released DriveSeg Open data set of .

By publishing DriveSeg, MIT and Toyota are working hard to promote the research of auto drive system , Just like human perception , Auto drive system regards the driving environment as a continuous flow of visual information .

so far , The autopilot data provided to the research community is mainly composed of a large number of static single images , These images can be used to recognize and track common objects found in and around roads , For example, bicycles , Pedestrian or traffic lights , By using “ Bounding box ”. by comparison ,DriveSeg Contains a more accurate pixel level representation of many common road objects , But it is through the lens of continuous video driving scene . This type of full scene segmentation is useful for identifying more amorphous objects that do not always have this definition and unified shape （ For example, road construction and vegetation ） Especially useful .

according to Sherony That's what I'm saying , The data flow provided by video based driving scene perception is more similar to dynamic , Real world driving . It also enables researchers to explore data patterns over time , This may drive machine learning , Progress in scene understanding and behavior prediction .

DriveSeg It's free , Researchers and academia can use it for non-commercial purposes . The data consists of two parts （manual And semi-auto） form .DriveSeg（manual） It was caught during the day on the busy streets of Cambridge, Massachusetts 2 branch 47 Seconds of high-resolution video . Of this video 5,000 Frame usage 12 Human labels per pixel of road like objects are densely annotated .

DriveSeg（semi-auto） It's from MIT Advanced vehicle technology （AVT） Extracted from Alliance data 20,100 Video frames （67 individual 10 Second video clip ）.DriveSeg（Semi-auto） And DriveSeg（manual） Have the same pixel level semantic annotation , Except that the annotation is through MIT Developed a new semi-automatic annotation method . Compared with manual annotation , This method takes advantage of manual and computational efforts , Rough annotation of data can be done more effectively at a lower cost . The purpose of creating this data set is to evaluate the feasibility of annotating various real driving scenes , And the evaluation is based on AI The marking system creates pixel markers on the potential of training vehicle perception systems .

Dataset home page ：https://agelab.mit.edu/driveseg

Link to the original text ：http://news.mit.edu/2020/mit-toyota-release-visual-open-data-accelerate-autonomous-driving-research-0618

The good news ！

Xiaobai learns visual knowledge about the planet

Open to the outside world

 download 1：OpenCV-Contrib Chinese version of extension module 

 stay 「 Xiaobai studies vision 」 Official account back office reply ： Extension module Chinese course , You can download the first copy of the whole network OpenCV Extension module tutorial Chinese version , Cover expansion module installation 、SFM Algorithm 、 Stereo vision 、 Target tracking 、 Biological vision 、 Super resolution processing and other more than 20 chapters .


 download 2：Python Visual combat project 52 speak 
 stay 「 Xiaobai studies vision 」 Official account back office reply ：Python Visual combat project , You can download, including image segmentation 、 Mask detection 、 Lane line detection 、 Vehicle count 、 Add Eyeliner 、 License plate recognition 、 Character recognition 、 Emotional tests 、 Text content extraction 、 Face recognition, etc 31 A visual combat project , Help fast school computer vision .


 download 3：OpenCV Actual project 20 speak 
 stay 「 Xiaobai studies vision 」 Official account back office reply ：OpenCV Actual project 20 speak , You can download the 20 Based on OpenCV Realization 20 A real project , Realization OpenCV Learn advanced .


 Communication group 

 Welcome to join the official account reader group to communicate with your colleagues , There are SLAM、 3 d visual 、 sensor 、 Autopilot 、 Computational photography 、 testing 、 Division 、 distinguish 、 Medical imaging 、GAN、 Wechat groups such as algorithm competition （ It will be subdivided gradually in the future ）, Please scan the following micro signal clustering , remarks ：” nickname + School / company + Research direction “, for example ：” Zhang San  +  Shanghai Jiaotong University  +  Vision SLAM“. Please note... According to the format , Otherwise, it will not pass . After successful addition, they will be invited to relevant wechat groups according to the research direction . Please do not send ads in the group , Or you'll be invited out , Thanks for your understanding ~